共查询到18条相似文献，搜索用时 140 毫秒
1.
逐步门限自回归模型及其建模方案 总被引：2，自引：0，他引：2
针对门限自回归模型在模型识别方面的不足，提出了逐步门限自回归模型，并同时给出了该模型的一种建模方案。数值实例表明，逐步门限自回归模型在模拟和预报稳定上比一般门限自回归模型有一定程度的提高。 相似文献
2.
3.
多元门限回归模型的一种建模方法 总被引：12，自引：0，他引：12
本文根据门限自回归模型的基本思想[1]，提出一种多元门限回归模型的建模方法。其特点是充分考虑了预报系统中某些特殊预报因子突变点对预报关系的改变作用。数值实例表明，该模型在模拟和预报精度上比一般线性逐步回归模型有一定程度的提高。 相似文献
4.
南宁市空气臭氧污染的TAR模型预报研究 总被引：1，自引：0，他引：1
以门限自回归统计方法为基础进行南宁城市臭氧日最大小时浓度TAR模型预报研究.选取日平均气压作为参与预报气象因子、180日作为建模资料长度、开方形式的数据转换后建立模型预报效果更佳.预报结果检验表明预报方案是可行的,并提出了改进建议. 相似文献
5.
6.
门限混合回归模型是由门限自回归模型发展而来。门限自回归模型(Threshold Autore-gressive Model)是由汤家豪于1978年首先提出的，该模型经过国内十几年的研究已经有了很大进展和完善。本文介绍门限混合回归模型在辽宁省一些地区季降水和温度预报中的应用。 相似文献
7.
非线性门限自回归模型用于时间序列的外推预报 总被引：3，自引：0，他引：3
对一些呈非线性变化的时间序列，如果勉强用线性统计模型来描述，效果往往不理想。本文利用非线性的自激励门限自回归模型（ＳＥＴＡＲ）、开环门限自回归模型（ＴＡＲＳＯ）对我省记录年代较长的烟台年降水量序列进行建模分析，并探讨分析了模型的稳定性，最后利用稳定的ＳＥＴＡＲ模型进行外推预报。 相似文献
8.
1 引言 门限混合回归模型是由门限自回归模型发展而来。门限自回归模型(Threshold Autoregressive Model)是由汤家豪于1978年首先提出的,该模型经过国内十几年的研究已经有了很大进展和完善。本文介绍门限混合回归模型在辽宁省一些地区季降水和温度预报中的应用。 相似文献
9.
10.
基于SSA-AR方法的MJO指数预报模型试验 总被引：7，自引：0，他引：7
采用奇异谱分析(SSA)与自回归向量(AR)预报模型相结合的方法,对热带地区大气季节内振荡(MJO)指数向量作自适应滤波意义下的预报试验。结果表明,通过对MJO原始序列进行SSA的分解重建,无论采用对重建的分量序列进行AR(P)建模的方案,还是利用对重建合成序列进行AR(P)建模的方案,均可得到两周以上的MJO指数预报能力,其提前20天指数预报值与实况之间平均相关系数达到0.5,与直接对MJO原始序列进行AR建模相比较,该方法有较高的预报技巧和超前预报能力,预报效果也较稳定,故将SSA-AR方案进一步完善,可望作为MJO指数业务预报的有效模型。 相似文献
11.
In order to achieve the best predictive effect of the Partial Least Squares (PLS) regression model, Particle Swarm Optimization (PSO) algorithm is applied to automatically filter the optimal subset of a set of candidate factors of PLS regression model in this study. An improved version of the Particle Swarm Optimization-Partial Least Squares (PSO-PLS) regression model is applied to the station data of precipitation in Southwest China during flood season. Using the PSO-PLS regression method, the prediction of flood season precipitation in Southwest China has been studied. By introducing the precipitation period series of the mean generating function (MGF) extension as an alternative factor, the MGF improved PSO-PLS regression model was also build up to improve the prediction results. Randomly selected 10%, 20%, 30% of the modeling samples were used as a test trial; random cross validation was conducted on the MGF improved PSO-PLS regression model. The results show that the accuracy of PSO-PLS regression model and the MGF improved PSO-PLS regression model are better than that of the traditional PLS regression model. The training results of the three prediction models with regard to the regional and single station precipitation are considerable, whereas the forecast results indicate that the PSO-PLS regression method and the MGF improved PSO-PLS regression method are much better than the traditional PLS regression method. The MGF improved PSO-PLS regression model has the best forecast performance on precipitation anomaly during the flood season in the southwest of China among three models. The average precipitation (PS score) of 36 stations is 74.7. With the increase of the number of modeling samples, the PS score remained stable. This shows that the PSO algorithm is objective and stable. The MGF improved PSO-PLS regression prediction model is also showed to have good prediction stability and ability. 相似文献
12.
Summary Rainfall anomaly patterns are obtained for the city of Barcelona from a statistical and a spectral point of view. The time
series consists of monthly rainfall amounts recorded over 128 years without interruption. Monthly positive and negative anomalies,
obtained as the difference between monthly amounts and monthly threshold values, are used for both types of analyses. The
threshold levels are derived form the deciles of theoretical monthly rainfall distributions, which have been previously modelled
by the gamma distribution. Positive and negative anomalies of the monthly rain amounts are investigated for these threshold
levels. The statistical analysis is applied to each decile considered, yielding empirical exponential laws that can be used
to forecast the cumulative number of episodes of consecutive months with either positive or negative anomalies equalling or
exceeding a fixed length. A set of linear laws, relating the expected rainfall amount cumulated during an episode of a fixed
length, is also deduced. It is worthy of mention that, independently of the decile considered, all the exponential and linear
laws have satisfactory regression coefficients. At the same time, it has also been possible to establish the evolution of
the coefficients of these laws with respect to the different deciles considered. The exponential laws for episodes of positive
and negative anomalies are the starting point, together with two hypotheses, to model probabilities of repeated long episodes
over an arbitrary number of years and their return periods in terms of the Poisson distribution model. Moreover, power spectra
are derived for anomalies relative to the 50% decile at monthly and seasonal scale. The spectral estimates obtained are then
compared with theoretical spectra deduced from possible Markovian or random behaviour of the time series of anomalies. Finally,
the significant spectral peaks are discussed and compared with other significant spectral components deduced for some areas
of the Mediterranean domain.
Received November 11, 1999 Revised February 28, 2000 相似文献
13.
I. Matyasovszky 《Theoretical and Applied Climatology》2003,74(1-2):69-75
Summary ?Homogenized monthly and annual mean temperatures for ten locations in Hungary from 1901 to 1999 are analyzed. A principal
component analysis was performed and the first new component containing 94.5% of the total variance has been retained. A linear
regression of this variable on a sea level pressure NAO index results in relatively weak correlations. In order to consider
the trends in both data series, a polynomial of years is added to the regression. After a selection of the optimal polynomial
orders by Akaike’s criteria the correlation coefficients are significantly increased. The Southern Oscillation Index (SO index)
characterizing the El Ni?o – Southern Oscillation (ENSO) is then incorporated in the relationship via a nonlinear, threshold
model. The threshold model consists of the above linear regressions but is conditioned on the SO index threshold variable.
The rationale behind this approach is to allow a change of model performance according to ENSO phase. The thresholds are not
pre-specified but are estimated from the data, while the number of thresholds is chosen by Akaike’s criteria. A likelihood
ratio test shows an improvement of these models over the linear model with very strong significance levels, except in June.
Received March 25, 2002; revised June 20, 2002; accepted June 23, 2002 相似文献
14.
基于数字高程模型(digital elevation model,DEM)高程数据和D8算法实现了湖南中小河流流域子流域划分。利用洪水预报网络传播模型,构建中小河流各子流域径流拓扑结构并提出了空间子流域等效面雨量的概念。基于等效面雨量序列和径流观测序列,构建了基于时空概念的隐马尔可夫降水-径流模型,并利用该模型计算了中小河流各子流域不同时间尺度条件下暴雨致灾临界面雨量阈值,最后利用2010—2015年实况资料对阈值进行检验,结果证明相比传统统计法,新方法的计算结果与传统方法结果一致且具有很好的准确性和稳定性。基于2015年6月的一次暴雨灾害预报,证明了该方法适应于业务化运行。 相似文献
15.
Illia Horenko 《Dynamics of Atmospheres and Oceans》2010,49(2-3):164-187
A method for clustering of multidimensional non-stationary meteorological time series is presented. The approach is based on optimization of the regularized averaged clustering functional describing the quality of data representation in terms of several regression models and a metastable hidden process switching between them. Proposed numerical clustering algorithm is based on application of the finite element method (FEM) to the problem of non-stationary time series analysis. The main advantage of the presented algorithm compared to Hidden Markov Models (HMMs) and to finite mixture models is that no a priori assumptions about the probability model for the hidden and observed processes (e.g., Markovianity or stationarity) are necessary for the proposed method. Another attractive numerical feature of the discussed algorithm is the possibility to choose the optimal number of metastable clusters and a natural opportunity to control the fuzziness of the resulting decomposition a posteriory, based on the statistical distinguishability of the resulting persistent cluster states. The resulting FEM-K-trends algorithm is compared with some standard fuzzy clustering methods on toy model examples and on analysis of multidimensional historical temperature data locally in Europe and on the global temperature data set. 相似文献
16.
利用门限回归模型和非线性滑动回归模型,分析天津市2014-2019年日用电量和气温之间的非线性关系,并计算得到不同响应关系下的阈值气温。结果表明：天津地区的阈值气温在不同响应关系下存在明显差异,在"V"型非线性响应关系下,阈值气温为18.8℃,在"U"型响应关系下,线性不对称模型的舒适区范围为12.3-23.4℃,非线性模型的舒适区范围为13.7-21.7℃;对比不同模型的预测效果,认为"U"型优于"V"型模型,非线性模型优于线性模型。对阈值气温的影响要素分析表明,相对湿度对舒适区与冷却区的阈值气温影响较大,阈值气温在相对湿度为30%-50%时较相对湿度为50%-70%时偏大2.2℃,相比之下,相对湿度对舒适区与加热区的阈值气温影响不大;天津地区阈值气温会随时间发生明显变化,2002-2005年的舒适区范围较2014-2019年偏大1.4℃。在实践应用中,应根据模型的需求,并充分考虑相对湿度和时间变化的影响选择阈值气温,从而提升用电量预测的准确率。 相似文献
17.
基于EMD 和集合预报技术的气候预测方法 总被引：3，自引：0，他引：3
气候系统是典型的非平稳性系统,然而对于气候观测数据的处理通常是在时间序列平稳的假定下完成的,比如气温和降水的多步预报,这通常会导致预报准确度较低。为改进该缺陷,首先将非平稳数据序列分解成平稳的、多尺度特征的本征模态函数分量(IMF),再使用数值集合预报与逐步回归分析相结合的方式对每一个IMF 分量构建不同的预报模型,最后线性拟合成预报结果。通过Visual Studio 2008 开发平台使用上述方法建立了一个短期气候预报系统,采用广西区88 个气象站1957—2005 年的2 月距平气温数据进行实际验证。结果表明,相对于普通预测和单一预测方法,加入了EMD 和集合预报技术的方法在仅用历史资料进行多步预测的情况下,对于气候的变化趋势以及突发性气候具有更好的预报能力。 相似文献
18.
针对目前采用的统计方法存在的不足, 即在选择预报因子时没有考虑预报因子之间的相关性, 挑选的预报因子由于非正交, 使回归计算的结果不稳定, 给计算带来一定的误差。该文提出把一元线性回归分析、自然正交函数 (EOF) 和逐步回归方法结合起来, 从而得到一种新的建立统计预报模型的方法。以西安市采暖期和夏季SO2日均浓度为预报对象, 使用该方法建立预报模型。拟合及预报试验表明, 这些预报模型不但可以很好地拟合变化趋势, 而且还能作出较准确的预报, 采暖期预报的级别命中率为72.5 %, 夏季级别预报命中率为100%。通过对比试验, 此方法优于目前常用的逐步回归方法, 具有很好的应用前景。 相似文献